
THE ODIN SYSTEM
CAN SEARCH AND AI ACTUALLY READ YOUR SITE?
Odin scores a site out of 100 across ten categories — fifty points for search readiness, fifty for AI visibility, because we sell those separately and price them the same.
Every number is computed from what it observes: your robots.txt, your sitemap, your structured data, which AI crawlers you allow, whether your content exists before JavaScript runs. Run it twice and you get the same score. Anything we cannot measure from outside is left out of the calculation rather than counted against you.
THE METHODOLOGY, PUBLISHED
TEN CATEGORIES. TWO HALVES. NO NUMBER A MODEL MADE UP.
The score is arithmetic over checks that either passed, partly passed or failed. A language model writes the explanation afterwards and is given the findings verbatim — it cannot introduce one we did not observe.
Crawlability
Whether a crawler can reach and index the site at all.
Mobile & Speed
Real-world field data where Google publishes it — and an honest “not measured” where it does not.
On-Page & Structure
Titles, headings and internal links: whether the page states what it is.
Local Signals
Name, address, phone, and the markup that ties them together.
Structured Data
Not a count of schema blocks — whether they are valid, complete and agree with the visible page.
Entity Clarity
Can a machine tell which business this is and reconcile it with the wider web.
AI Crawler Access
Whether GPTBot, ClaudeBot, PerplexityBot and the rest are allowed to read you.
llms.txt
An emerging convention, weighted low on purpose. Checked because a wrong one actively hurts.
Answer-Shaped Content
Is there a paragraph a model could lift whole and attribute to you.
Citation-Worthiness
Proxies only, and labelled as proxies on the report.
TWO THINGS WE WILL NOT DO
We will not score what we could not measure. Most small-business sites have no real-world speed data, because Google only publishes it above a traffic threshold. Counting that as a failure would knock seven points off nearly every small site for a reason that is not a defect — so those points leave the calculation entirely and the report says so. If we could measure less than 70% of the checklist, no overall score is published at all.
We will not invent a finding. Every deduction carries the evidence that produced it — the literal line from your robots.txt, the count, the URL. If there is no evidence, there is no finding.
One more, stated because the rest of the report depends on it: blocking AI crawlers is a legitimate business decision. If yours is deliberate, ignore that section — it is reported as a choice, not as a mistake.

GO BEYOND THE SCORE.
Odin reads what it can from outside. On a 20-minute call we go through the findings with you and tell you which half — search readiness or AI visibility — is actually your constraint.
Book a Strategy Call